Volume 343, Issue 3 · 10 articles from 3 issues
OpenText introduces a data access governance maturity model for assessing unstructured data risks as AI tools roll out.
OpenText notes that 80% or more of stored enterprise information is unstructured, file-based content ranging from redundant junk to an organization's crown jewels.
Data Access Governance identifies these files, evaluates their sensitivity and relevance and secures them, and AI deployments such as Microsoft Copilot are now its biggest driver.
This first of three parts introduces the DAG Maturity Model for assessing data access risks.
OpenText explains how validating and connecting engineering documents to asset records makes them AI-ready.
Engineering organizations hold critical knowledge in P&IDs, CAD drawings, data sheets, maintenance records and procedures scattered across repositories, often with mismatched asset tags and poor metadata.
OpenText argues this information becomes AI-ready when asset data is extracted and validated, documents are linked to authoritative asset records, relationships are discoverable and governed processes route, review and approve changes when discrepancies arise.
The post describes combining Cad-Capture's Asset Data Intelligence with OpenText content management.
OpenText's DAG maturity series covers Level 3 enforcement and Level 4 certified governance of data access.
The second part of OpenText's data access governance series moves beyond Level 1 unmanaged risk and Level 2 visibility.
It covers Level 3, enforcing appropriate access through policies, automation and ongoing permission management, and Level 4, adding business-owner certifications, formal access reviews and documented accountability.
OpenText says the goal is to replace slow manual remediation with automated, policy-driven defenses, as set out in its newly published DAG Maturity Model.
OpenText argues a modern Dynamics 365 integration strategy is the data foundation for AI-ready operations.
Timed to Community Summit NA 2026, this OpenText post argues AI does not fix disconnected processes and fragmented data and can expose gaps faster.
Organizations that invested in Dynamics 365, ERP modernization and automation often accumulate a mix of EDI connections, APIs, custom integrations and spreadsheets.
OpenText says a modern Dynamics 365 integration strategy reduces complexity, improves visibility and creates the trusted data foundation needed for automation and AI.
OpenText Core Capture 26.4 adds built-in integration with the Capture Aviator AI add-on for complex tables, handwriting and poor scans.
OpenText outlines the October 2026 release of OpenText Core Capture 26.4, part of its SaaS intelligent document processing family that also includes versions for Salesforce and SAP.
The release adds built-in integration with the new AI document understanding add-on, OpenText Capture Aviator.
OpenText says this improves data accuracy and expands straight-through processing of complex tables, handwriting and low-quality scans.
It also extends automation to new or frequently changing document types and provides LLM extraction directly through OpenText to avoid AI licensing complexity.
OpenText brings Capture Aviator, an LLM-powered document understanding add-on, to OpenText Capture and Core Capture.
OpenText says capture workflows handle most documents well, but handwritten forms, overlapping tables and low-quality scans still require significant human effort.
It cites research that 37% of employees see inadequate automation and repetitive tasks as a key obstacle.
To close the gap, OpenText announced Capture Aviator in June 2026 as an AI document understanding add-on.
The add-on uses LLM-powered visual and language understanding to read documents the way people do, interpreting context and layout to extract data from unstructured content.
OpenText warns that forgotten files such as old exports and HR spreadsheets become a real risk once AI tools can surface them.
OpenText says protecting sensitive data from shadow AI requires first knowing what data exists and what is sensitive.
Many organizations hold forgotten files such as old migration exports, HR spreadsheets saved to team shares and folders from long-closed deals.
These files were once relatively harmless because they were hard to find.
OpenText argues AI changes that, because sanctioned and unsanctioned AI tools can search, summarize and surface this content.
OpenText argues knowledge-driven GenAI grounded in company data beats generic LLMs for high-stakes customer communications.
OpenText says GenAI improves customer experience only when it draws on what a business actually knows.
A model that has never seen a company's policies, claims history or customer records can produce polished but wrong letters.
Many enterprises have tried generic AI writing tools only to spend time checking figures, fixing tone and confirming regulatory compliance.
The post makes the case for knowledge-driven GenAI, especially for teams sending high-stakes customer communications.
OpenText argues that blocking unsanctioned AI tools rarely works and that shadow AI risk should be managed by protecting the data.
OpenText notes that employees routinely paste customer emails and spreadsheets into AI assistants simply to get their work done.
It defines shadow AI as the use of AI tools, models or features without the knowledge or approval of IT and security teams.
Examples include chatbots on personal accounts, AI features in SaaS apps and browser extensions with built-in assistants.
OpenText argues that finding and blocking every unsanctioned tool rarely solves the problem, because the real risk is the data flowing through them.
OpenText concludes its shadow AI series by arguing for data-centric security that protects sensitive data even in unapproved AI tools.
OpenText says shadow AI reaches data organizations may not know exists or know is sensitive.
After discovering sensitive data and clearing out redundant, obsolete and trivial data, the next step is protecting the data that remains.
Customer records, payment data and health information keep multiplying as data is copied into analytics platforms, shared with partners and pulled into AI workflows.
In this final post of its shadow AI series, OpenText explains how data-centric security protects that data even when copies land in unapproved AI tools.